Introduction to Data Analytics
Ziyuan Huang
Last Updated: November 13, 2025
An Example of what to Expect in Descriptive Analytics: Ex.1.1
library(datasets)
data("sunspot.month") # Load embedded dataset
head(sunspot.month, 10)
## Jan Feb Mar Apr May Jun Jul Aug Sep Oct
## 1749 96.7 104.3 116.7 92.8 141.7 139.2 158.0 110.5 126.5 125.8
An Example of what to Expect in Descriptive Analytics: Ex.1.1
- Visualize the trend of the dependent variable from the
dataset:
- Here, will visualize the data to better understand the sunspots
historical trend from 1749 to 2013.
- In a few weeks, we will cover how to make these plots!
library(ggplot2)
library(dplyr)
# Convert to tibble with time index
sunspot_df <- tibble(
time = seq_along(sunspot.month),
sunspots = as.numeric(sunspot.month)
)
ggplot(sunspot_df, aes(x = time, y = sunspots)) +
geom_point(alpha = 0.5, color = "steelblue") +
labs(
y = "Number of Sunspots",
x = "Time (Months since 1749)",
title = "Historical Sunspot Activity"
) +
theme_minimal() +
theme(plot.title = element_text(hjust = 0.5))
An Example of what to Expect in Predictive Analytics: Ex.2.1
- Let’s import some data from the quantmod package.
- Provides access to both stock data, but also useful functions for
stock analysis.
- Understand the Data Structure and Data Types:
library(quantmod)
library(lubridate)
# Define date range (last 5 years)
start_date <- Sys.Date() - years(5)
end_date <- Sys.Date() - days(2)
# Fetch stock data with error handling
suppressWarnings(
getSymbols("AMZN", src = "yahoo", from = start_date, to = end_date, auto.assign = TRUE)
)
## [1] "AMZN"
## An xts object on 2020-11-13 / 2025-11-10 containing:
## Data: double [1253, 6]
## Columns: AMZN.Open, AMZN.High, AMZN.Low, AMZN.Close, AMZN.Volume ... with 1 more column
## Index: Date [1253] (TZ: "UTC")
## xts Attributes:
## $ src : chr "yahoo"
## $ updated: POSIXct[1:1], format: "2025-11-13 17:47:41"
An Example of what to Expect in Predictive Analytics: Ex.2.1
# Split data for training (first 1199 observations)
train_size <- min(1199, nrow(AMZN))
train_data <- AMZN[1:train_size, ]
# Build predictive model
predictive_model <- lm(
formula = AMZN.Close ~ AMZN.High + AMZN.Low + AMZN.Volume,
data = train_data
)
summary(predictive_model)
##
## Call:
## lm(formula = AMZN.Close ~ AMZN.High + AMZN.Low + AMZN.Volume,
## data = train_data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -6.9513 -0.8685 -0.0221 0.9510 9.3134
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -1.445e-02 2.484e-01 -0.058 0.954
## AMZN.High 5.268e-01 2.427e-02 21.701 <2e-16 ***
## AMZN.Low 4.732e-01 2.473e-02 19.137 <2e-16 ***
## AMZN.Volume -8.230e-10 1.845e-09 -0.446 0.656
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 1.415 on 1195 degrees of freedom
## Multiple R-squared: 0.9985, Adjusted R-squared: 0.9985
## F-statistic: 2.704e+05 on 3 and 1195 DF, p-value: < 2.2e-16
An Example of what to Expect in Predictive Analytics: Ex.2.1
- Visualize the Models Performance:
# Diagnostic plots for model assessment
par(mfrow = c(2, 3), mar = c(4, 4, 2, 1))
plot(predictive_model, which = 1, main = "Residuals vs Fitted")
plot(predictive_model, which = 2, main = "Q-Q Plot")
plot(predictive_model, which = 3, main = "Scale-Location")
plot(predictive_model, which = 4, main = "Cook's Distance")
plot(predictive_model, which = 5, main = "Residuals vs Leverage")
par(mfrow = c(1, 1)) # Reset layout
An Example of what to Expect Analytics: Ex.2.1
# Visualize predictions
plot(prediction,
type = "l",
col = "steelblue",
lwd = 2,
main = "Predicted Amazon Closing Prices",
xlab = "Time Index",
ylab = "Predicted Close Price ($)",
las = 1)
grid(col = "gray90")
